NovAST: Novel cell type discovery and Annotation in Spatial Transcriptomics

This directory contains the source code for the following study: NovAST: A Deep Learning Framework for Automated Label Transfer and Novel Cell Type Discovery in Spatial Transcriptomics.

Python Dependencies

The file named requirements.txt contains the required python pacakges.

Web Application

The example code is prototyped and documented in the Jupyter Notebook.

Installation Guide

Here is the sample command to install packages into a python environment:

# clone the code from GitHub
git clone https://github.com/YMa-lab/NovAST.git
# create a new environemnbt with python==3.11
python -m venv NovAST_env
# activate the python environemtn 
source NovAST_env/bin/activate
# install the package
pip install .

Tutorial

For a step-by-step tutorial, see the documentation: NovAST Docs.